Bandura Flashcards
Aim
To investigate the extent to which children would imitate aggression modelled by an adult
To investigate the effects of gender on imitation.
Sample
72 children
36 females and 36 males
Aged between 37-69 months (3yrs 1month and 5 yrs 9 months), mean age 52 months.
All from Stanford University Nursery School (USA)
Independent Variables (I.V.s)
There were three IVs in this experiment:
Behaviour of the model: In the aggressive condition the model attacked the Bobo doll and verbally abused it; in the non-aggressive condition the model played quietly and ignored the Bobo doll; Bandura also had a CONTROL CONDITION where there was no model at all.
Sex of the model: half of the children saw a male model, the other half a female one
Sex of the child: half the children were boys, half were girls
Dependent Variables (D.V.s)
The two observers gave each child an “aggression score” by recording the following categories:
Imitative aggression: This is aggression that directly copies the model. It could be physical (for example, punching the doll in the nose) , verbal (for example, repeating the phrases “Pow!” or “Sock him in the nose“)or non-aggressive verbal (saying things like “He keeps coming back for more”)
Partially-imitative aggression: This includes using the mallet on other toys or sitting on Bobo.
Non-imitative aggression: This includes punching Bobo in new ways, saying hostile things not said by the model and using the dart gun aggressively
Method / Design
Method - Structured Observation
Experimental design- Matched pairs design - on pre-existing aggressiveness using 5 point scales by the experimenter and their nursery school teacher.
They did this by observing the children in the nursery and judged their previous aggressive behaviour on four 5-point rating scales. It was then possible to match the children in each group so that they had similar levels of aggression in their everyday behaviour.
Condition
Condition 1
24 observe aggressive behaviour in models
Condition 2
24 Children observe non-aggressive behaviour in models
Control group
24 children did not observe models
Procedure
Phase 1: The observation of the aggressive/non-aggressive model.
The children were sat in one corner of the room and settled by the experimenter who then left when the model entered the room.
The model sat in the opposite corner of the room to the child.
In the models behaved in a standardised way e.g.
aggressive model condition - An example of physical aggression was “raised the Bobo doll and pommeled it on the head with a mallet”, An example of verbal aggression was, “Pow!” and “Sock him in the nose”.
Phase 2: The mild aggression arousal.
Experimenter told the child the toys were the experimenter’s very best toys and she had decided to reserve them for the other children
Phase 3: The observation of the participants- the children
The experimenter had originally planned to leave this room however, due to a number of children getting upset when the experimenter tried to leave some even refusing to stay. Therefore the researcher had to stay in the room.
Data was recorded for 20 minutes every 5 seconds (time sampling) therefore resulting in a total of 240 recordings for each child.
BANDURA’S QUALITATIVE DATA
The children were shocked by the aggressive female model, with one boy saying“That ain’t no way for a lady to behave!”and a girl saying“That girl… was acting just like a man!”
Controls
The model behaved in a standardised pre-scripted way:.
The aggressive model would: spend he first minute playing quietly then the rest of the time attacking the Bobo doll. The aggressive routine included putting the doll on its side, sitting on it, repeatedly punching it on the nose, hitting it on the head with the mallet, throwing it in the air and kicking it round the room. The model would say specific things like: “Hit him down!”, “Pow!” or “He keeps coming back for more!“.
Conclusions
Witnessing aggression in a model can be enough to produce aggression by an observer.
Children selectively imitate gender-specific behaviour
Key Theme
‘External influences on children’s behaviour
WHO
Bandura studied 72 children, 36 females and 36 males, aged between 37-69 months (mean age 52 months) and all from Stanford University Nursery School (USA).
WHAT
This study investigated the extent to which children would imitate aggression modelled by an adult.
HOW
This was done by exposing the children to an aggressive model, non-aggressive model or no model and then observing the children for 20 minutes to see whether their level of aggression was higher or lower.
FINDINGS
It was found that children in the aggressive model condition showed significantly more imitation of physical and verbal aggressive behaviour than children in the non-aggressive model or control (no model) conditions.
Method -Strength
P A strength of a structured observation is that it provides an objective and reliabile way of collecting data.
E This type of observation provides the researcher with an observation schedule, a set time to observe and behavioural categories they must tally when they observe. This reduces the likelihood that personal opinions or predictions will interfere with data collection and increases the likelihood that someone else could observe the same behaviour and get the same results.
E In phase three of the Bandura study the researcher and the male model observed the children for 20 minutes, tallying down every 5 seconds if they saw any of the following behaviours; imitative aggression, partially imitative aggression or non-imitative aggression. Both observers obtained similar data and therefore had inter-rater reliability. This suggests that data collection was both objective in its collection and reliable.
Method - Weakness
P A weakness of a structured observation is it can be said to be low in internal validity.
E This means the researcher may not be measuring what they set out to mention. This method only collects quantitative data and therefore only tells the reader the number of times a behaviour occurs, It does not explain the context or reasons for behaviours. This leads to a lack of detail in terms of behaviours occurred and the social context.
E In Bandura’s study we know that boys showed a higher number of aggressive behaviours than girls. However, it does not give detail as to how violent each act was.
Data - Strength
P A strength of quantitative data is that it is high in reliability.
E Numeric data is objective in its analysis, this means that it will not be affected by the thoughts, opinions or experiences of the researcher and therefore another researcher can repeat the analysis and gain the same results.
E In the Bandura study an aggressive score was calculated for each child by adding up the number of aggressive acts (imitative/ partially imitative and non-imitative aggression) that were seen in phase three. This calculation should remain the same regardless of who does the analysis.
Data - Weakness
P A weakness of quantitative data is that it can be said to be low in internal validity
E As numeric data is limited in its detail it can often be unclear why the researchers got the results they did, as we are unable to see the full picture. Therefore, it can be hard to draw conclusions, as we cannot be sure that the IV causes the change in the DV.
E In Banduras study we only know that the children in the aggressive model condition showed more aggression. We cannot be sure that this was due to the exposure to the model. E.g. they could of seen other children being aggressive with the Bobo doll in the past and copied their behaviour or have this toy at home and previously played with it.
Validity - Strength
P A strength of the Bandura et al’s study is that it is high in internal validity.
E A number of controls have been put in place to ensure that extraneous variables do not impact the DV.
E Bandura used a one-way mirror in phase three to observe the children. The children would have been unable to see the experimenter and the male model watching and making notes on their behaviour. This may control for socially desirable behaviours of the children, the children acting in a way that makes them look good in the eyes of the observer e.g. playing nicely with the non-aggressive toys.
Validity - Weakness
P A weakness of the Bandura et al’s study is that it can be said to be low in external validity, specifically ecological validity.
E This means that the results from this study may not be generalisable beyond the setting in which they were found. This is because the procedure does not reflect something which is likely to occur in a real-life setting.
E Bandura’s study takes children out of their natural setting to be studied in laboratory conditions. Within the study they are also in quick succession taking into three different rooms, the first of which they are exposed to a stranger acting aggressive towards a bobo doll. This is highly unlikely to be something they would have been exposed to if they were not taking part in this study.,
Reliability - Strength
P A strength of the Bandura et al’s study is that it is high in external reliability.
E The procedure was standardised meaning that a number of controls have been put in place to ensure that all participants had a similar experience. Therefore, another researcher should be able to repeat the study and receive the same results.
E In Bandura’s study the models the children were exposed to in phase one of the experiment acted in a pre-scripted way. E.g. The aggressive model would: spend he first minute playing quietly then the rest of the time attacking the Bobo doll. The aggressive routine included putting the doll on its side, sitting on it, repeatedly punching it on the nose, hitting it on the head with the mallet, throwing it in the air and kicking it round the room. The model would say specific things like: “Hit him down!”, “Pow!” or “He keeps coming back for more!
Reliability - Weakness
P A weakness of the Bandura study is that it could be said to be low in external reliability.
E This means that some parts of the procedure may have been different for each participant and therefore making it harder to repeat the procedure exactly, reducing the chance that the same results could be found.
E In Bandura’s study phase 3 the models were given a script to follow. However, due to individual differences between the male and female model may have acted differently e.g. due to one saying ‘pow’ in a more aggressive manner or hitting the bobo doll harder with the mallet. These variations could also occur for each model as it would be hard to repeat the performance exactly each time. This means it would be hard to replicate the study exactly how it was done originally
Sampling Bias
Whole Procedure
